The landscape of copy number variations in <scp>F</scp>innish families with autism spectrum disorders

Rare de novo and inherited copy number variations (CNVs) have been implicated in autism spectrum disorder (ASD) risk. However, the genetic underpinnings of ASD remain unknown in more than 80% of cases. Therefore, identification of novel candidate genes and corroboration of known candidate genes may broaden the horizons of determining genetic risk alleles, and subsequent development of diagnostic testing. Here, using genotyping arrays, we characterized the genetic architecture of rare CNVs (<1% frequency) in a Finnish case-control dataset. Unsurprisingly, ASD cases harbored a significant excess of rare, large (>1 Mb) CNVs and rare, exonic CNVs. The exonic rare de novo CNV rate (∼22.5%) seemed higher compared to previous reports. We identified several CNVs in well-known ASD regions including GSTM1-5, DISC1, FHIT, RBFOX1, CHRNA7, 15q11.2, 15q13.2-q13.3, 17q12, and 22q11.21. Additionally, several novel candidate genes (BDKRB1, BDKRB2, AP2M1, SPTA1, PTH1R, CYP2E1, PLCD3, F2RL1, UQCRC2, LILRB3, RPS9, and COL11A2) were identified through gene prioritization. The majority of these genes belong to neuroactive ligand-receptor interaction pathways, and calcium signaling pathways, thus suggesting that a subset of these novel candidate genes may contribute to ASD risk. Furthermore, several metabolic pathways like caffeine metabolism, drug metabolism, retinol metabolism, and calcium-signaling pathway were found to be affected by the rare exonic ASD CNVs. Additionally, biological processes such as bradykinin receptor activity, endoderm formation and development, and oxidoreductase activity were enriched among the rare exonic ASD CNVs. Overall, our findings may add data about new genes and pathways that contribute to the genetic architecture of ASD.